Source-primed Multi-turn Conversation Helps Large Language Models Translate Documents
This addresses document-level translation quality for users of large language models, though it is incremental as it builds on existing multi-turn approaches.
The paper tackles omission errors in document-level machine translation by LLMs, proposing a source-primed multi-turn conversational method that decomposes documents into segments and iteratively translates them while reusing context; it outperforms single-turn and independent segment translation in multiple automatic metrics.
LLMs have paved the way for truly simple document-level machine translation, but challenges such as omission errors remain. In this paper, we study a simple method for handling document-level machine translation, by leveraging previous contexts in a multi-turn conversational manner. Specifically, by decomposing documents into segments and iteratively translating them while maintaining previous turns, this method ensures coherent translations without additional training, and can fully re-use the KV cache of previous turns thus minimizing computational overhead. We further propose a `source-primed' method that first provides the whole source document before multi-turn translation. We empirically show this multi-turn method outperforms both translating entire documents in a single turn and translating each segment independently according to multiple automatic metrics in representative LLMs, establishing a strong baseline for document-level translation using LLMs.